Bryn Mawr Classical Review

Martti Leiwo, Hilla Halla-aho, Marja Vierros (ed.), Variation and Change in Greek and Latin. Papers and monographs of the Finnish Institute at Athens, 17. Helsinki: Foundation of the Finnish Institute at Athens, 2012. Pp. iii, 177. ISBN 9789526721149.

Reviewed by Jerker Blomqvist, Lund University (jerker.blomqvist@gmail.com)

Version at BMCR home site

[The Table of contents is listed at the end of the review.]

"Variation is the essence of language", Martti Leiwo reminds us in the introductory chapter of the volume under review, and variation is a prerequisite of linguistic change. Competing linguistic forms always exist in a given speech community; when one of them starts to dominate over the others—or, alternately, disappears—the effect is a change in the linguistic system. The book contains the revised versions of papers presented at a colloquium arranged by its editors at the Finnish Institute of Athens in September 2009 with the aim to study what kinds of variation existed within the definable varieties (or "registers" or "fractions") of Latin and ancient Greek and to determine, if possible, the reasons for the occurring variations. The period chosen for study extends from the fourth century B.C. to the sixth A.D. Linguistic data are retrieved from a variety of texts, mostly from original documents (papyri, graffiti, formal inscriptions) but also from literary works preserved in medieval manuscripts.

Variation within language communities has been studied with modern linguistic methods at least since the 1960s. For modern languages there now exist large digitized text corpora with grammatical annotation. In them, variation can be studied with predominantly quantitative methods. The ancient corpus languages mostly offer only a limited amount of data for the study of an individual phenomenon and require a different approach. Leiwo and the other contributors repeatedly underline that a qualitative method of analysis is necessary. All textual evidence should be taken into account, but the significance of individual data must be assessed with reference to the purpose of the text, the situation in which it was produced, and the writer's person.

The available corpus of Greek and Latin texts is constantly expanding, if only by small steps. Leiwo demonstrates, with a syntactical phenomenon as example, how new data compel us to adjust our ideas about the degree of variation existing in the ancient languages. Mandilaras, in 1973, had claimed that the construction with τοῦ + infinitive had become regular with expressions like μὴ ἀμελήσῃς 'don't forget (to)'. Leiwo shows that recently published ostraca papyrus letters display a wide variation of constructions with μὴ ἀμελήσῃς (vel sim.): infinitive without article, aorist imperative, aorist subjunctive, future indicative, ἵνα clause, etc. The syntactic variation was much greater than indicated by the material available to Mandilaras.

The relevance of a qualitative evaluation of the data is demonstrated by Trevor Evans. He has worked with the vast Zenon Archive (third century B.C.), applying "a combined linguistic, onomastic, and paleographic method". With paleographic analysis he identifies three individuals who sent letters to Zenon and wrote some of them with their own hand. One of them, on the whole, writes standard Greek, i.e. the variety of the Attic dialect that became the norm of formal prose writing in the Hellenistic period. Frequent deviations from the orthographic norm characterize a second writer. In the third set of texts too syntactical errors are usual, presumably due to bilingual interference, for the writer's name, Petosiris, may indicate Egyptian ethnicity (although names are not reliable as indicators of ethnicity, as other contributors to the volume point out). Other letters from the same senders are written by professional scribes in their service, and these show individual stylistic features. Thus, these texts reflect a linguistic community characterized by a high degree of variation. Evans' findings could probably be generalized for Hellenized Egypt as a whole, possibly for other Greek diasporas in the Mediterranean area as well.

Marja Vierros investigates another set of Greek texts from Egypt, contracts for the sale of landed property in the town of Pathyris from the second and first centuries B.C. The texts show that the writers had problems with case syntax and frequently confused the genitive and the accusative in the long nominal phrases by which the purchased properties were identified. Vierros aims to decide whether those deviations from the norm are due the idiolects of individual, presumably bilingual, scribes or if they belonged to a variety of the language in general use among the staff of the agoranomos. She concludes that the documents were written by Egyptians who were used to drawing up analogous contracts in their own language. The contracts in Greek were modeled on their Egyptian counterparts and by a "phraseological transfer" a number of Egyptianisms crept into the Greek documents. The degree of confusion exhibited by individual scribes varies, perhaps due to varying efficiency of training, but on the whole the deviations from the Greek norm are not idiolectic phenomena but features common to the linguistic community of the agoranomos office.

The sociolinguistic landscapes of both Latin and ancient Greek had one fateful feature in common: they were both dominated by a normative, standard variety of the language, which was taught at schools and regarded as the only correct language, both in antiquity and afterwards. Classicists have tended to regard that norm as the only genuine manifestation of the language and all deviations from it as marks of linguistic incompetence. In her paper Hilla Halla- aho shows that certain features that have been classified as manifestations of "vulgar" Latin are attested in texts of the pre-classical period (Terentius); they cannot be examples of later corruption or degeneration of Classical Latin, and we may suspect that this also applies to a number of other phenomena attested in post-classical texts. As a linguistic term "Latin", without further qualification, traditionally refers to Classical Latin, and "vulgar Latin" is everything that cannot be attested in the canonical texts. Halla-aho recommends that the qualification "vulgar" should be avoided for such phenomena, that "Latin" should be used as the generic term for all varieties of the language, and that Classical Latin should be regarded as one of those varieties, on par with the others.

The existence of a normative, prestigious variety of the language tends to retard the natural development of the colloquial language, but at the same time colloquial forms will creep into texts intended by their writers to follow a literary standard. Gerd V.M. Haverling shows how the tensions between the conservative and colloquial varieties of Latin manifest themselves in the manuscript tradition of a great number of literary texts of Late Antiquity and the Middle Ages. The examples discussed include all sorts of linguistic phenomena: reduction of the case system, case forms versus preposition phrases, subordination and its alternatives, modifications of the verbal system, vocabulary. The individual manuscripts that preserve a certain text often deviate in these respects. Thus, Haverling can present a vast amount of data but must admit that it is mostly impossible to decide whether the non-standard phenomena appearing in a manuscript are due to the original writer or to the scribe who copied the text, and a scribe who knew his Latin could have "corrected" sub-standard forms appearing in the original. Only in the few cases when a writer reproduces, more or less verbally, an actual or imaginary conversation do such texts afford reliable testimony to the development of colloquial Latin.

Giovanbattista Galdi treats a morphological phenomenon of sub-standard Latin, viz. the nominative plurals in as of first-declension substantives (e.g. filias for filiae), which occur in, mostly non-literary, texts from the first century B.C. until early Middle Ages (and are likely to have survived into most modern Romance languages). The phenomenon has been discussed repeatedly for the last 100 years, as appears from Galdi's overview of previous research. His own survey of the material shows that, even when 35 uncertain cases are deducted, the number of known instances has almost doubled since 1909, and examples are now known from all the western provinces of the empire, plus Dacia, the two Moesiae and the two Pannoniae. The phenomenon was certainly widespread in spoken, sub- standard Latin. Galdi does not claim to provide a universally valid explanation of the phenomenon but points to three factors that are likely to have favored its progress: (i) the general spreading of the accusative to new syntactical functions; (ii) a tendency, after the merger of ae with e, to reintroduce the characteristic a vowel in all forms of first-declension substantives; (iii) in tomb inscriptions, where the as forms are common, it was felt important to distinguish a female dedicatee, whose name was a dative in ae or e, from, e.g., her dedicating daughters, who could be referred to, in the nominative, as filias.

Rolando Ferri describes Roman strategies for giving negative answers to appeals and demands. Ferri takes most of his material from literary texts, comedies in the first place and the grammarians' metalinguistic comments on them. The focus is on discursive devices rather than on words and phrases. There seems to be little difference between standard and sub-standard language varieties in this field. Rather it is pragmatic factors—the mutual relationship of the parties involved and the immediate situation—that decide what strategy is chosen for making the answer at the same time polite, non-offensive and unambiguous.

Three papers treat the influence of Greek and Latin on each other. Paolo Poccetti also includes other Italic languages in his survey of originally Greek divine and mythological names in documents from Italy, and it appears that more than one Greek dialect has left traces in his material. The name of the hero Aias becomes Aivas in Etruscan, so its model must have come from a Greek dialect with preserved intervocalic digamma, whereas Latin Aiax, gen. Aiacis, stems from an Ionic dialect without digamma but may testify to the existence of a Greek stem in κ instead of ντ , unattested but possibly traceable in the names of Aias' grandfather Αἰακός and his descendants, the Αἰακίδαι. Ionic Ἀπόλλων becomes Apollo in Latin, and also Etruscan, Marsian and Faliscan have o or u in the second syllable, whereas Oscan Appellunei and Vestinian Apellune mirrors Doric Ἀπέλλων. For Greek Ἄρτεμις/Ἄρταμις, Etruscan Artumes, Messapian Artemes, etc., the Romans substitute their own Diana. Greek influence is evident in these names but it may have several sources and use different routes, and all details are not clear.

Heikki Solin discusses the linguistic situation in Campania. At some date Latin must have become the dominant language in this area, where it had earlier been rivaled by other Italic languages and by Greek. Unfortunately, the epigraphic material does not reveal that date, for it is mostly later than this language change. Only Naples provides relevant material but, according to Solin, the permanence of Greek in this originally Greek colony is not typical of the situation in Campania. The Pompeian inscriptions are, with few exceptions, in Latin, but the graffiti contain more Greek. The interpretation of these texts is difficult, and Solin suggests that the standard editions should be corrected or reinterpreted in a considerable number of cases. He also points out that a graffito in Greek, strictly speaking, only shows that its author used Greek for writing graffiti, not how often he used the language for other purposes. Also onomastics and writing habits are unreliable indicators; a Greek name does not necessarily reveal a person's ethnicity or language habits, and writing Latin with Greek letters, or vice versa, perhaps only belongs to the genre conventions of graffiti writing. Greek was present in Pompeii, but to what extent cannot be determined.

Eleanor Dickey explains why the Latin loanwords in Greek are essential for the more general question of Latin influence on Greek. She criticizes the existing corpora for indiscriminately classifying any Latin word appearing in a Greek text as a loanword. Only words integrated in the system of the borrowing language—phonetically, orthographically, morphologically, semantically—and appearing with some frequency tell us something about the impact of the source language. Most Latin loanwords have been recorded from papyri and inscriptions, but Dickey finds the testimony of those texts problematic and focuses on certain types of literary texts. When a Greek lexicographer explains an uncommon Greek word with its Latin-derived equivalent, when a purist like Phrynichus condemns the current use of a Latin word, or when Latin words are used without further explanation in the New Testament and other texts intended solely for speakers of Greek, the Latin words qualify as integrated loanwords. Dickey has created a corpus of more than 600 such words (which she modestly calls "preliminary"). Her material includes "a solid group of real loanwords" that had been fully integrated with Greek. They are not confined to texts with specialized contents but appear in practically all semantic areas and, contrary to expectation, a majority of them were not borrowed in late Antiquity, after the imperial capital had been established on Greek soil, but more than half of them had been introduced, if not fully integrated, into Greek before the fourth century A.D.

Dickey's list of 600 fully integrated words does not appear in the volume. One gets the impression that considerations of space have prevented also other contributors from developing their arguments in full. Several of them refer the reader to their own more comprehensive treatments of the theme of their papers, others present their material more fully. Bibliographies, some of them vast, appear with the individual papers. Indexes of passages and words would have been useful but are not added. The phenomena of variation and change have not always been fully understood by classicists, which could be deplored, since the two classical languages, with their long, documented history provide valuable material for the study of these matters. The present volume is a stimulating reminder of that fact.

Table of Contents

Preface
Abbreviations
Martti Leiwo, 'Introduction: Variation with Multiple Faces'
Hilla Halla-aho, 'What Does 'Latin' Mean? A Terminological Pamphlet'
T.V. Evans, 'Linguistic and Stylistic Variation in the Zenon Archive'
Marja Vierros, 'Phraseological Variation in the Agoranomic Contracts from Pathyris'
Eleanor Dickey, 'Latin Loanwords in Greek: A Preliminary Analysis'
Paolo Poccetti, 'Reflexes of Variations in Latin and Greek through neither Latin nor Greek Documentation: Names of Greek Religion and Mythology in the Languages of Ancient Italy'
Heikki Solin, 'On the Use of Greek in Campania'
Rolando Ferri, 'How to say No in Latin: Negative Turns, Politeness and Pragmatic Variation'
Giovanbattista Galdi, 'Again on as-nominatives: A New Approach to the Problem'
Gerd V.M. Haverling, 'Literary Late Latin and the Development of the Spoken Language'
List of Contributors